It is proposed to examine the mechanism of enzyme-catalyzed and nonenzymic reactions of the coenzyme thiamin (vitamin B1). The long-term objective of the proposed research is to determine how thiamin-dependent enzymes stabilize or avoid such unstable carbanion "intermediates" as the thiazolium Cs-ylide and the C alpha-carbanion/enamine that have been implicated in a number of enzyme mechanisms; a particular interest is the role, if any, of the intrinsic binding energy of the substrate(s) or coenzyme in stabilizing the carbanions. The immediate, short-term goals are to understand the mechanism and catalysis of two nonenzymic reactions involving these moderately unstable carbanions and to determine whether or not these carbanions exist as discrete intermediates on a thiamin-dependent enzyme. The research will focus on the mechanism and catalysis of enzymic and nonenzymic aldol-type addition reactions involving the thiazolium Ca-ylide, and the extent to which the mechanisms of these addition reactions are determined by the lifetime of the ylide-type carbanion "intermediate". Using standard kinetic methods and isotopic probes, the mechanism of a model aldol-type addition reaction between thiazolium ylides and aldehydes in aqueous solution will be determined and compared with the mechanism of thiamin-dependent decarboxylation of alpha-keto acids by pyruvate decarboxylase (PDC). The structure of the transition state for nonenzymic addition of the ylide to aldehydes will be characterized, and the pa of PDC-bound thiamin C2-H will be estimated. Secondly, the mechanism for nonenzymic C alpha-proton transfer from thiazolium carbinols to form the C alpha-carbanion/enamine in aqueous solution will be examined using standard kinetic methods and isotopic probes to determine the role of general base catalysis, inductive effects, and internal return. This information will be used to estimate the pa of the C alpha-proton, determine whether the C alpha-carbanion has a significant lifetime, estimate the intrinsic barrier for proton transfer from these iminium ion-activated carbon acids, and determine the effect of stereoelectronic control (a requirement for proper orbital overlap) on the rate and intrinsic barrier for abstraction of the C alpha- proton in aqueous solution. Using isotopic probes, it will be determined whether the C alpha-carbanion has a significant lifetime on PDC.